study of a gas turbine cycle with hydrogen combustion

I

Bachelor’s Thesis

Study of a gas turbine cycle with hydrogen

combustion

under the supervision of

Willinger, Reinhard; Ao.Univ.Prof. Dipl.-Ing. Dr.techn.

E302 – Institute of Energy Systems and Thermodynamics

submitted to the Faculty of Mechanical and Industrial Engineering

of Technische Universität Wien

for the degree of Bachelor of Science

by

Álvaro de la Peña Pinteño

Matr.Nr. 12045101

Kandlgasse 30, 1070 Wien

II

Statutory Declaration

This thesis is the result of my own work and includes nothing that is the outcome

of work done in collaboration except as specified in the text.

It is not substantially the same as any that I have submitted, or, is being

concurrently submitted for a degree or diploma or other qualification at Technische

Universität Wien or any other University or similar institution except as specified in the

text. I further state that no substantial part of my thesis has already been submitted,

or, is being concurrently submitted for any such degree, diploma or other qualification

at Technische Universität Wien or any other University or similar institution except as

specified in the text.

Vienna, June 2021

…………………………………...

Álvaro de la Peña Pinteño

III

Acknowledgements

I would like to express my huge gratitude to the professor Reinhard

Willinger, from the Institute of Energy Systems and Thermodynamics from the

Technische Universität Wien for his help during the development of this thesis,

providing investigations and advise when it was required. Furthermore, I would

also like to highlight the support of my family and friends during these years of

education.

IV

Abstract

As it is well known, electricity is an issue which plays an essential role in

any society. However, the impacts its production provokes to the environment

must be taken into consideration, mainly those due to greenhouse gases

produced by burning carbon-containing compounds.

In this situation, a possible way of reducing these gases would consist in

the production of electricity using hydrogen, which could also be produced by

renewable sources.

Following this idea, this bachelor thesis will seek, through the compilation of

different articles, research and previous work, to offer an analysis on how a pre-

existing gas turbine cycle which works using natural gas as fuel could be

implemented and transformed, to work with hydrogen.

Firstly, the properties of both fuels (natural gas and hydrogen) will be

compared and assessed. Then, the effects that this change, in terms of fuel,

causes in the turbo machinery of the cycle will be evaluated. Subsequently, I will

discuss the settings required with the intention of minimizing possible losses in

terms of performance or cycle efficiency.

After this point, this research will focus on another very important element

in the face of this change: the combustion chamber. To begin, a comparison of

the reactions that take place in it will be carried out, also explaining the

particularities that the use of hydrogen as fuel has associated, later a comparison

will be made between the different combustion systems that can be used, with

the purpose to determine which will be the most appropriate for this element.

To conclude this research, a classification of the different, most prominent

types of hydrogen which can be found today will be provided. Finally, some of the

techniques which can be used to produce this hydrogen will be discussed.

V

Kurzfassung

Elektrizität ist bekanntlich ein Thema, das in jeder Gesellschaft eine

wesentliche Rolle spielt. Allerdings müssen die Auswirkungen seiner Produktion

auf die Umwelt berücksichtigt werden, die hauptsächlich auf die Treibhausgase

zurückzuführen sind, die durch die Verbrennung kohlenstoffhaltiger

Verbindungen entstehen.

Ein möglicher Weg, diese Gase zu reduzieren, bestünde in dieser Situation

in der Stromerzeugung mit Wasserstoff, der auch aus erneuerbaren Quellen

gewonnen werden könnte.

Dieser Idee folgend wird diese Bachelorarbeit versuchen, durch die

Zusammenstellung verschiedener Artikel, Forschungen und bisheriger Arbeiten

eine Analyse zu bieten, wie ein bereits bestehender Gasturbinenkreislauf, der mit

Erdgas als Brennstoff arbeitet, implementiert und transformiert werden könnte,

um zu funktionieren mit Wasserstoff.

Zunächst werden die Eigenschaften der beiden Gase (Erdgas und

Wasserstoff) verglichen und bewertet, dann die Auswirkungen, die diese

Änderung in Bezug auf den Brennstoff in der Turbomaschiner des Kreislaufs

verursacht, bewertet und anschließend die erforderlichen Einstellungen mit den

Absicht, mögliche Verluste in Bezug auf Leistung oder Zykluseffizienz zu

minimieren.

Danach wird sich diese Forschung angesichts dieser Veränderung auf ein

weiteres sehr wichtiges Element konzentrieren: die Brennkammer. Zunächst wird

ein Vergleich der darin ablaufenden Reaktionen durchgeführt, wobei auch die

Besonderheiten der Verwendung von Wasserstoff als Brennstoff erläutert

werden, später ein Vergleich zwischen den verschiedenen einsetzbaren

Verbrennungssystemen mit den Zweck zu bestimmen, welches für dieses

Element am besten geeignet ist.

Zum Abschluss dieser Forschung wird eine Klassifizierung zwischen den

verschiedenen bekanntesten Arten von Wasserstoff bereitgestellt, die heute

gefunden werden können, und schließlich werden einige der Techniken

diskutiert, mit denen wir diesen Wasserstoff herstellen können.

VI

Resumen

Como es bien sabido, la electricidad actualmente es un elemento que juega

un papel fundamental en cualquier sociedad, sin embargo, deben ser tenidos en

cuenta los impactos que su producción generan sobre el medio ambiente,

principalmente aquellos debidos a los gases de efecto invernadero que se

producen al quemar compuestos que contienen carbono.

Ante esta situación, una posible medida que podría suponer una gran

reducción de estos consistiría en la producción de esta electricidad usando

hidrógeno, que a su vez podría ser producido con fuentes renovables.

Siguiendo esta línea de investigación, este trabajo de fin de grado

pretenderá, mediante la recopilación de distintos artículos, investigaciones y

trabajos previos, ofrecer un análisis sobre cómo podría implementarse y

transformar un ciclo de turbina de gas preexistente, que funciona empleando gas

natural como combustible, a uno que use hidrógeno.

En primer lugar, se realiza una comparación entre las propiedades de

ambos elementos (hidrogeno y gas natural), posteriormente se evaluarán los

efectos que este cambio en cuanto al combustible provoca en la turbo

maquinaría del ciclo, debatiendo posteriormente los ajustes a realizar con la

intención de minimizar las posibles pérdidas en cuando a las prestaciones o la

eficiencia del ciclo.

Tras este punto, esta investigación se centrará en otro elemento muy

destacado frente a este cambio: la cámara de combustión. Para comenzar, sobre

ella se realizará una comparativa de las reacciones que allí se producen, y sobre

las particularidades que el uso de hidrogeno como combustible tiene asociado,

posteriormente se realizará una comparativa entre los distintos sistemas de

combustión que pueden ser empleados, con la finalidad de determinar cuál será

el más apropiado para este elemento.

Para concluir esta investigación, se ofrece una breve clasificación con los

distintos tipos de hidrógenos más destacados que se pueden encontrar

actualmente para finalmente comentar posibles métodos a partir de los cuales

puede ser producido este elemento.

VII

Resum

Com és ben sabut, l'electricitat actualment és un element que juga un paper

fonamental en qualsevol societat, no obstant això, han de ser tinguts en compte

els impactes que la seua producció generen sobre el medi ambient, principalment

aquells deguts als gasos d'efecte d'hivernacle que es produeixen en cremar

compostos que contenen carboni.

Davant aquesta situació, una possible mesura que podria suposar una

gran reducció d'aquests consistiria en la producció d'aquesta electricitat usant

hidrogen, que al seu torn podria ser produït amb fonts renovables.

Seguint aquesta línia d'investigació, aquest treball de fi de grau pretendrà,

mitjançant la recopilació de diferents articles, investigacions i treballs previs,

oferir una anàlisi sobre com podria implementar-se i transformar un cicle de

turbina de gas preexistent, que funciona emprant gas natural com a combustible,

a un que use hidrogen.

En primer lloc, es realitza una comparació entre les propietats de tots dos

combustibles(hidrogen i gas natural), posteriorment s'avaluaran els efectes que

aquest canvi quant al combustible provoca en la turbo maquinària del cicle,

debatent posteriorment els ajustos a realitzar amb la intenció de minimitzar les

possibles pèrdues en quan a les prestacions o l'eficiència del cicle.

Després d'aquest punt, aquesta investigació es centrarà en un altre

element molt destacat enfront d'aquest canvi: la cambra de combustió. Per a

començar, sobre ella es realitzarà una comparativa de les reaccions que allí es

produeixen, i sobre les particularitats que l'ús d'hidrogen com a combustible té

associat, posteriorment es realitzarà una comparativa entre els diferents

sistemes de combustió que poden ser emprats, amb la finalitat de determinar

quin serà el més apropiat per a aquest element.

Per a concloure aquesta investigació, s'ofereix una breu classificació amb

els diferents tipus d'hidrògens més destacats que es poden trobar actualment

per a finalment comentar possibles mètodes a partir dels quals pot ser produït

aquest element.

Table of contents

VIII

Table of contents

Statutory Declaration ................................................................................. II

Acknowledgements ................................................................................... III

Abstract ..................................................................................................... IV

Kurzfassung ............................................................................................ V

Resumen ............................................................................................... VI

Resum .................................................................................................. VII

Table of contents .................................................................................... VIII

Nomenclature............................................................................................ IX

1.- Introduction ........................................................................................... 1

2.- Moving from natural gas to hydrogen ................................................... 5

2.1 Hydrogen compared to natural gas. .................................................. 5

2.2 Effects on turbomachinery ................................................................ 7

3.- Process of combustion ....................................................................... 13

3.1 Chemical reaction and implications ................................................ 13

3.2 Diffusive compared to premixed combustion .................................. 14

3.3 Performance and emissions ........................................................... 16

3.4 Premixed lean direct injection combustion ...................................... 21

4.- Where could hydrogen be obtained from? .......................................... 25

4.1 Classification of hydrogen ............................................................... 25

4.2 Methods to produce hydrogen ........................................................ 26

5.- Conclusion .......................................................................................... 29

Figures and tables ................................................................................... 31

Tables ................................................................................................... 31

Figures .................................................................................................. 31

Bibliography ............................................................................................. 33

Appendix .................................................................................................. 35

Nomenclature

IX

Nomenclature

Abbreviation Description

NG Natural gas

NOx Nitric oxides

Wc Work of the compressor

SFT Stoichiometric flame temperature

LDI Lean direct injection

https://dictionary.cambridge.org/dictionary/english/oxide

Study of a gas turbine cycle with hydrogen combustion

1

1.- Introduction

Nowadays, energy plays an essential role in modern societies, it could be

considered as an essential issue for economic growth and the improvement of

our well-being. It is the angular stone of any industrial, manufacturing or

production process, being fundamental for the development of any region or

county. Consequently, whenever a country undergoes a process of economic

growth, it will be alongside an increase in energy consumption, as is shown in

Figure 1.

Therefore, worldwide energy consumption and its production has been

constantly and gradually rising, except for a lower growth in the years 2009 and

2020. These two years experienced a decrease in economic growth due to an

economical and sanitary crisis, respectively, as can be seen in Figure 1.

This mentioned energy comes from a variety of sources, we can distinguish

two opposite groups: on the one hand we have the non-renewable resources,

their use and exploitation provoke polluting emissions, they are mainly those

denominated as fossil fuels (coal, oil and natural gas). On the other hand, we can

find renewable sources. These have a low impact for the environment, causing

less pollution. Their main disadvantage is their dependence on nature’s

behaviour. Within this group, solar, wind and hydroelectric energy stand out.

These systems cannot be used for the whole day due to the lack of sunrays at

night or the noncontinuous of the wind.

However, the facilities that use high-capacity power systems demand

energy during the whole day, to face this situation we can find two choices. The

first choice consists of the use of an energy storage system like batteries.

Figure 1 Global primary energy consumption by source.(Iea Org, 2021). Figure 1 Global primary energy consumption by source.(Iea Org, 2021).


2

However, batteries are very expensive and have significant adverse effects on

the environment due to the use of highly polluting mines inside. The other

alternative would be the employment of a renewable energy source which could

be used with independence of the climatology agents. That could be the case of

hydrogen, produced via renewable energy sources and used as fuel.(Koç et al.,

2020).

As seen in the Figure 1, traditional biomass is an energy source that has

remained practically constant since the XIX century. From the middle of the same

century and motivated by the second industrial revolution, we can observe the

development of coal and oil as energy sources, being at this moment the most

important sources of energy, providing us with approximately 55% of the energy

consumed in the year 2019, they are used as fuel in vehicles and in many heavy-

duty electricity productions plants. Beginning in the 50s, a quick development in

the use of natural gas as source of energy takes place. This gas is mainly used

in heavy-duty turbines as a fuel, being at this moment one of the most important

production system of electricity to supply cities and countries. To conclude the

analysis of this plot, in the last year we can observe the increase of the use of

renewable sources as energy sources. This is a fundamental point to reduce our

pollution levels and end the emission of greenhouse gases.

This thesis will focus on the changes required to use these power plants

with hydrogen as fuel, what will mean a giant step in the reduction of global

pollution, being the right direction to achieve the Paris agreement’s goals with the

aim to reverse the current climate change situation.

This natural gas is used, in many cases, as fuel to produce electricity in a

gas turbine cycle, those are based on the Rankine cycle. The simplest consists

in four steps (illustrated in Figure 2),

firstly the air enters the cycle at

ambient pressure and temperature.

Then it will go through the

compressor, it can have different

shapes mainly axial (if the air flow is

parallel to the axis of the compressor)

but also the flow can be radial (if the

flow has a certain angle with the axis).

We use work (Wc) to increase the

pressure of the gas, ideally this

process should be an isentropic compression (1->2s). Depending on the

efficiency of the compressor, the compression will more closely resemble an

isentropic or not.

Figure 2 Representation of a Brayton cycle with its more significant points.(ARREGLE et al., 2002)


3

After this step, the air flow goes to the combustion chamber, there, at

constant pressure, the combustion takes place (in Figure 1 shown as line 2->3),

these exhausted gases go to the turbine, where we can find an expansion in

different steps depending on the number of row blades of the turbine, from there

we get the energy to supply the demand.

This is the simplest cycle, there are, nonetheless many variations. We can,

for instance, include a regenerator to improve the global efficiency, two or three

steps in the process of compression, some turbines with different steps of

combustion can also be combined. In any case, they are derivatives of the

Brayton cycle described previously and have the purpose to reduce the energy

needed to contribute to the cycle and maximize the energy we get.

One of the most outstanding characteristics of the gas turbine cycle is the

easy regulation it provides. We can quickly change the electricity it supplies

according to the variations in the demand. It should also be mentioned that it has

less elements than the steam power cycle and it is smaller. Thus it offers us a

wide variety of implementation: we can find from small turbines called

microturbines which are able to generate from 20 to 350 kW to heavy industrial

turbines able to supply up to 350 MW and with an efficiency around 40%,

including as well those used in the aerospace industry for the transport of goods

and passengers, they have less power but a higher efficiency (up to 45%) due to

a higher development and optimization of the material used.

NG is mainly composed of methane (CH4) and nitrogen. During its

combustion, it releases substances to the environment, when those emissions

interact within an ecosystem, they demand some source of material or energy

resource within said ecosystem; some of these sources are water, oxygen, solar

energy and biological systems as described by (Díaz & Ascencio, 2009). Despite

the greenhouse emissions due to its content in carbon, NG also produces other

contaminants during its combustion. When increasing the combustion

temperature, the process of nitrogen oxides formation will be more likely (since

an increase in temperature reaction is reflected in an increase in the constant

equilibrium of the NOx formation process). This increase of NO formation will grow

parabolic with temperature. On the other hand, NO2 formation presents the same

trend, (however, the amount of NO2 produced it will be almost 10,000 times less

than the NO produced) (Díaz & Ascencio, 2009).

The last contaminant to have into consideration is SOx, its production

depends on the entrance temperature of the gases in the turbine SOx are

unstable and react quickly with water in the air producing hyposulphurous acid.

This causes an acid deposition known as acid rain, which can seriously affect the


4

plant cover, they are also the cause of the degradation of a wide range of

materials.

After this brief introduction about the consumption, supply and generation

of energy, this bachelor thesis will address and discuss some topics related to

this aspect of the engineering, specially related to the transformation of energy

from the primary sources so it can be used for the human needs. Firstly, the

benefits and disadvantages of using a heavy-duty gas turbine with hydrogen as

fuel or with natural gas will be shown, paying special attention to the

performances of each kind of machine and the environmental impact each one

may have. Then will be explained if an existing heavy-duty gas turbine could be

used with hydrogen combustion, explaining if would be completely incompatible,

or otherwise, the requirements or modifications any component would need.

Another aspect to deal with will be the combustion chamber, it is one of the

more outstanding components of the power cycle in this transformation. The

changes it will face to be able to work with hydrogen will be discussed, explaining

the chemical reaction which will take place there, the size or shape it would have

and the different methods we will implement to minimize the production of NOx

gases. The next point of this thesis will be focussed on the benefits of hydrogen,

such as the reduction of pollution it is associated with. The main part of this thesis

will end with a brief discussion and comparation between some of the actual

methods that can be used to get the hydrogen.

This research will conclude with a summary of the arguments and issues

previously covered and trying to foretell the next steps into the development and

implementation of this technology.


5

2.- Moving from natural gas to hydrogen

2.1 Hydrogen compared to natural gas.

When we want to change the fuel supply of a turbine power cycle from

natural gas to hydrogen, should be taken and assessed the differences of each

compound, evaluating their potential benefits or disadvantages, should also be

analysed the performances of the whole cycle focusing also on each of its

components.

Firstly, the differences between natural gas and hydrogen must the

discussed. On the one hand, NG is composed of different elements but it is

methane in approximately 95% (UnionGas, 2017). On the other hand, hydrogen

is one of the simplest elements, just integrated by one proton and one electron.

It is really abundant on earth but mostly mixed with oxygen or carbon, not in its

pure form, what means we cannot use it directly to our needs, it requires its

production and purification (some methods will be briefly explained at the end of

this research). Naturally, those processes require money and energy. However,

it is also a storable element. Once it is produced, it can be stored at a low pressure

as many other similar gases used for industrial processes being ready to use

when it is necessary, it can also be transported by hydrogen pipelines (similar to

NG) or even by train. One of the properties which stands out from hydrogen is

the high energy it contains per mass, with a value of approximately 120MJ/kg

compared to the NG (50 MJ/kg) however, it one of the lightest gases known so it

has a poor rate of energy per volume (10 MJ/Nm3).

When it burns, the flame is colorless. Due to its higher energy, hydrogen

can achieve higher temperatures. The approximately flame temperature for a mix

of 19.6% in volume of hydrogen is 2321K (Badía, 2005). The speed of

propagation of the flame, 2.65 m/s, being approximately six and a half time higher

than the speed of the methane, this favors a possible explosion, while diffusivity

and density tend to reduce its probability, especially in opened spaces. In closed

spaces, the hydrogen escape takes place with decreased temperature which

reduces the risk. On the other hand, chemical reaction proceeds with volume

reduction, so instead of an explosion what happens is an implosion (Gutierrez

Jodral, 1968). This issue will be addressed in the next chapter.

Other difference that must be said is the diffusion coefficient of both fuels.

This coefficient represents the ease with which each solute moves in a given

solvent, in this case air, for the methane it has a value of 0,18 cm2/s. However, in

the case of the hydrogen it reaches a value of 0,61 cm2/s (nearly three and a half


6

times higher). This property makes sense according to the flame speed

propagation discussed previously.

Hydrogen combustion is a topic which is nowadays under development and

investigation. Nevertheless, a limiting factor with those tests is the actual

availability of hydrogen. Tests lasting around an hour can burn through the

available stored hydrogen, which can take up to a week to replenish This is a

significant logistic challenge because large quantities of hydrogen gas are

needed, but there is not large-scale production yet (Filn, 2019). In the table which

follows we can see a summary of the differences between the technical properties

from both of fuels discussed.

Table 1 Comparation of properties between hydrogen and methane.

The most outstanding characteristic from hydrogen and the reason behind

its interest in the industrial processes is due to its combustion, which does not

produce any kind of greenhouse gases. Hence, it becomes really interesting

fuel to use to produce electricity, like in a gas turbine cycle, or even as fuel for

vehicles. The emissions that should be taken into consideration due to its high

temperature of the flame are the production of NOx. In the next chapter will be

discussed some methods to try to minimize them.

Property Hydrogen (𝐻2) Methane (𝐶𝐻4)

Calorific power (MJ/kg) 120 50

Autoignition temperature (ºC)

585 540

Adiabatic flame temperature in the air (ºC)

2,045 1,875

Ignition limit of ignition on air (% vol.)

4-75 5,3-15

Flame speed propagation (m/s)

2,65 0,4

Diffusion coefficient in air (cm2/s)

0,61 0,18

Toxicity No No


7

2.2 Effects on turbomachinery

Due to the differences with natural gas exposed previously, the use of

hydrogen combustion has many implications on the turbomachinery (compressor

and turbine). The combustion chamber will be discussed in the next chapter.

The most relevant implications that will be assessed and discussed are

mainly: a) variation of the enthalpy drop in the expansion, b) a variation of the

flow rate at the turbine inlet which, in turn, affects the turbine/compressor

matching, c) a variation of the heat-transfer coefficient on the outer side of the

turbine blades, affecting the cooling system performances (Chiesa et al., 2005)

d) the sealing of the system because the H2 molecules have a greater tendency

to leak than other gases because they are smaller so is easier to squeeze through

small cracks, gaps and other tolerances that cannot accommodate a larger CH4

molecule (Noble et al., 2021).

Combined with the hydrogen flow, we can find a specific proportion of steam

or N2, these substances are added with the aim to control the NOx production, as

I will be discussed in the next chapter.

As it can be seen in the chart of Figure 3, we can observe the influence of

hydrogen combustion, combined with steam in the isentropic enthalpy drop in

given conditions, compared to natural gas. The enthalpy drop in its minimum level

is around 5%, without the

presence of steam. This drop

is motivated firstly due to the

variation on the specific heat

and secondly because of the

variation of the temperature

drop in the expansion of the

gas in the turbine, as the chart

shows when we start adding

steam. This enthalpy drop

gradually increases reaching

around 13% when the diluent

to H2 mass ratio is 8,

simultaneously increasing the

mixture of steam carries to an

improvement in the specific

heat of the mix, causing a

decrease in temperature drop,

Figure 3 Variation of the stoichiometric flame temperature and of the inlet volume flow rate and isentropic enthalpy drop of a hydrogen and steam fueled gas turbine with respect to the reference natural gas case.(Chiesa et al., 2005).

.

Figure 2 Variation of the stoichiometric flame temperature and of the inlet volume flow rate and isentropic enthalpy drop of a hydrogen and nitrogen fueled gas turbine with respect to the reference natural gas case.(Chiesa et al., 2005). Figure 3 Variation of the stoichiometric flame temperature


8

this motivates an increase in the outlet temperature of the turbine. Lastly, it should

be mentioned that despite the lack of additional steam, the volume gas flow of

the hydrogen increases around 3% when compared to the natural gas.

Analysing Figure 4, we can observe some similitudes and differences. In

this case, the addition of nitrogen hardly causes a variation of the enthalpy drop

which remains constant at 5%.

Another fact to highlight is

related to the stoichiometric

flame temperature, while

adding steam we need a rate

of diluent over hydrogen mass

of around 3.5 to get a

temperature of nearly 2500K

in the case of nitrogen. This

rate increases up to 8, nearly

the double, due to the specific

heat of the steam nearly the

double of the nitrogen (1.04

J·g−1·K−1 for nitrogen and

2.01 for the steam).

The second implication

to be assessed would be the relation between the compressor and the turbine,

due to the variation of the fuel used and its changes into the flow rates, the

designed working point should be reevaluated to achieve the equilibrium between

both machines, according to research (Chiesa et al., 2005), we can find three

methods to fix this situation.

The first one keeps the compressor in the same working point as with

natural gas and reduces the inlet temperature to the turbine to match again, it will

provoke a reduction into the pressure ratio in the turbine, this solution would be

the most damaging for the cycle performances because the energy obtained from

a Brayton cycle is mainly based on the temperature of the flow at the entrance of

the turbine. Therefore, this system would not be implemented.

The second possible solution consists in leaving the variable guide vanes

of the compressor in the same angle and establish the same temperature at the

entrance of the turbine and regulate the gas flow adjusting the pressure ratio of

the compressor. This solution would increase the compressor and turbine

working ratio to reduce the mass flow and allow the matching between both

elements.

Figure 4 Variation of the stoichiometric flame temperature and of the inlet volume flow rate and isentropic enthalpy drop of a hydrogen and nitrogen fueled gas turbine with respect to the reference natural gas case.(Chiesa et al., 2005).

Table 3 Comparation of the results for each of the solutions implemented.(Chiesa et al., 2005).Figure 7 Variation of the stoichiometric flame temperature and of the inlet volume flow rate and isentropic enthalpy drop of a hydrogen and nitrogen fueled gas turbine with respect to the reference natural gas case.(Chiesa et al., 2005).


9

The third and last solution is based on leaving the temperature at the

entrance of the turbine and the pressure working ratio at the same point as they

were working previously with natural gas and on the other hand reduce the mass

flow by moving and adjusting the variable guide vanes of the compressor to

reduce the gas flow along the cycle.

It should also be mentioned that these three methods mentioned are

compatible and combinable but, as it has been said, the first one would not be

excessively interesting from an energetic point of view.

The third issue to be addressed according to (Chiesa et al., 2005) is the

cooling of the blades of the turbine. They are one of the most sensitive part of the

cycle, specially the first row, because they must deal with the mass flow when it

reaches its higher temperature just after the combustion chamber, so it is

essential to give them a good cooling. Blades can be considered as a crossflow

heat exchanges where the cooling system has to ensure that the temperature of

the material never reaches a value which can produce any damage to the

material.

Nowadays, many methods to accomplish this task can be found. Most of

these are based on an air flow which circulates by many channels inside the

blades. We can find some where the air circulates parallel to those channels or

some improvements, with systems where the air impacts with the inside face of

the blade or even some where the air circulates inside the blade but goes out by

many small holes, creating a protective wrap along the external face.

There are also few systems which use water at high pressure as refrigerant,

its main advantage is its higher specific heat so it will provide a better cooling to

the blades, but otherwise it supposes a huge increase into the weight of the

turbine.

In the aspect of cooling, changing to a hydrogen combustion has two

implications, on the one hand, the new elements of the mass flow have a higher

convective heat transfer coefficient. This implies a higher heat transfer between

the flow and the external face of the blades, hurting its thermal equilibrium. On

the other hand, the second implication, is motivated because of the increase into

the pressure ratio, this fact provokes the increase into the convective heat

transfer coefficient both in the external and internal faces of the blade because of

the fluid density increase, this effect decreases the cooling performances of the

air inside the blades.

To face these two consequences, we can distinguish between two solutions.

The first one could be to increase the cooling system flow. This would imply the


10

redesign of the blades of the turbine, this solution would not be recommended,

because the aim of this thesis is moving from a cycle based on natural gas

combustion to one with hydrogen with the minimum changes into the original

cycle and this would imply big changes for the design and manufacturing process

with large economic implications.

On the other hand, another solution could be suggested: to decrease the

temperature of the mass flow at the entrance of the turbine, even though it implies

the reduction of the power performances. Still, it seems to be the solution with

lower impact into the turbine original design.

In Table 2 we can find the results from the simulation carried out by Chiesa

et al. (2005). This table contains the results from the simulations in which the

three solutions discussed previously were tested. For every case, we can find the

results using just hydrogen or hydrogen combined with steam or nitrogen. On the

left hand, we can firstly find a column which correspond to the data from the

original turbine cycle based on natural gas combustion. It will be used as

reference to see its original performances.

Moving to the results of the first simulation (variating the variable guide

vanes and keeping the pressure ratio constant, in this case at 17), when the

hydrogen does not content any diluent, it provides by far the largest stoichiometric

flame temperature, with also a small increase in the power provided by the turbine

with respect to the natural gas. It also provides the highest efficiency rate

(58.32%).

However, due to the large flame temperature and the formation of

NOx emissions, we should mix steam or nitrogen with this hydrogen. With the

addition of steam, we have to reduce the air flow (to leave space for the addition

Table 2 Comparison of the results for each of the solutions implemented.(Chiesa et al., 2005).


11

of the steam) The improvement of some of the performances, mainly in the gas

turbine output, achieving 293 MW, can also be seen, motivated by the reduction

in the consumption of the compressor due to the lower air flow. There is, as well,

an increase in the enthalpy drop. Nonetheless, this does not imply a significant

change in the total output of the cycle. Lastly, it should be mentioned that the

efficiency decreases in nearly one point regarding the cycle with natural gas.

In the case of nitrogen, the inlet temperature is really similar to the case of

hydrogen. We can observe a larger reduction in the air flow, because, of the

higher diluent fuel mass ratio. Thus, we also need energy to compress this

nitrogen, the efficiency for this case remains practically constant as in the original

cycle.

The second column belongs to the result in which the compressor ratio is

increased to keep the air flow constant. We can observe similar results as in the

previous example. For this case, we need a higher amount of diluent, so the ratio

of nitrogen or steam is larger. The addition of nitrogen requires a higher pressure

than the steam, we can also observe a clear reduction in the inlet temperature,

probably as a result of the increase in the cooling mass flow. This decrease

provokes a reduction in the turbine performances and efficiency. Otherwise, the

power provided by the turbine is higher than previously, because the air flow

circulating along the turbine is around 8.5% higher in the case of steam and 15%

larger in the case of nitrogen as additive.

The last columns of the table show the results with a redesigned engine. In

general terms, for this case we obtain some improvements mainly with regard to

the power provided by the turbine and also a higher value for the efficiency

compared to the other cases but is not the aim of this research to redesign a new

engine.

According to the results provided by this simulation (Chiesa et al., 2005), in

which the implications of moving from a traditional NG to a hydrogen combustion

and their possible solutions to minimize its impact on the turbomachinery have

been assessed, we can conclude that from a technical view, it would be possible

to use a pre-existing gas turbine cycle and with the setting of some parameters,

use it with hydrogen.

Both solutions assessed would be appropriate for this task but would be

recommended the first solution: adjust of the variable guide vanes of the

compressor, to adequate the mass flow along the circuit allowing the matching

between the compressor and the turbine, combined with a reduction of the

temperature at the entrance of the turbine to preserve the health of specially the

first turbine blade row and avoid its damage. This solution would be preferred to


12

the second one, because it requires a lower impact on the original cycle, simply

the modification of the vanes, elements which are already included in the

compressor. Otherwise, the rise of the compressor ratio, proposed as solution

two, requires larger mechanical modifications as a larger compressor with a

bigger generator.

Despite the fact that in these cases we have obtained an efficiency of nearly

57%, it is because it has been analyzed with simplest Brayton cycle but including

some modifications as different steps for compression or expansion or reheat to

reduce the energy supplied by the chamber. This efficiency could be improved,

according to (Bannister et al., 1997), who also studied the use of hydrogen as

fuel, but in this case for a Rankine cycle based on water and steam and used for

the production of energy, which includes regeneration and reheat, it has provided

an efficiency of the cycle of around 71%, which is a really high value.

The last issue of the turbine to be taken into consideration is its sealing. At

high temperature and pressure ratio, increased leakage can become a serious

performance penalty, especially in the turbine section (Fadok & Diakunchak,

2008). This aspect can be easily fixed by using a special coating or a bondcoat

to prevent the leaks. This would be an easy and effective solution which would

not imply any penalty in the performances of the cycle.


13

3.- Process of combustion

3.1 Chemical reaction and implications

In both cases, the purpose is to obtain the energy storaged using oxygen to

produce its combustion. These are the formulations of the reactions:

CH4 + 202— > CO2 + 2H2O (1)

H2 +1

2O2— > H2O (2)

Obviously, both are exothermic reactions, but due to the characteristics of

each fuel they have some differences: the first one comes up by just observing

the formulation of the reactions. The combustion of hydrogen does not produce

any carbon emission, preserving the earth from greenhouse pollution.

The second difference, and among the most important one, is the high-

speed flame of hydrogen. It is nearly an order of magnitude higher than in the

case of the NG. In general terms, using the same conditions, a premixed, dry low

NOx nozzle would need 10 times higher flow velocity to prevent the flame from

flashing back and damaging the hardware (Noble et al., 2021).

The third difference is about the higher adiabatic flame that hydrogen

achieves in its combustion. It is somewhere in between 5 to 10% higher than

when using NG (Noble et al., 2021). This property has two implications: the first

one involves the increase in the potential production of NOxemissions (implication

which will be addressed in the next point) and the second one is about the

potential damaging or deterioration of the mechanic components of the

machinery. Another difference between these two combustions is the difference

in the flame geometry. Hydrogen flames are highly thermo-diffusively unstable,

creating corrugations and wrinkles along the flame that will grow, becoming

bigger. This is a fact that also increases its flame speed.

Theoretically, according to the higher heating value (120 MJ/kg for the

hydrogen and 50 MJ/kg for NG), the gas flow required for hydrogen to produce

the same energy would be nearly half than when compared to the case of NG.

Otherwise, hydrogen is, at the same conditions (temperature and pressure), eight

times less dense than the NG. Therefore, its volumetric lower heating value is

roughly a third of CH4 (Noble et al., 2021). This will provoke that the gas turbine

based on hydrogen combustion will require three times more volume than when

it was previously working with NG. A suggested way to solve this issue would be

by supplying H2 at a higher pressure. Nevertheless, it would imply a rise into the

operation costs of the cycle.


14

Combustion instabilities, also called as combustion dynamics, should be

considered as well. In the case of hydrogen combustion, the combustion

instability phenomenon has the challenging feature that its characteristics are

nonmonotonic with operating parameters (such as combustor inlet pressure,

combustor inlet temperature, fuel gas composition, etc.(Noble et al., 2021).

Consequently, it is difficult to predict the increment in combustion dynamics due

to the addition of H2, so different scenarios should be assessed.

These properties and differences discussed in this section related to the

combustion make a move from a traditional NG gas combustion to hydrogen

incredibly challenging. Therefore, some changes in the previous working

parameter to ensure the life of the machinery and obtain an appropriate

performance to make this process technically, energetically and economically

viable should be carried out.

3.2 Diffusive compared to premixed combustion

As a brief introduction, in the context of deflagrations, we can find two types.

Firstly, we can find the premixed flames. They are flames whose liberation heat

zone is located in the flame front. The temperature increases along the axial

position, starting in the bottom of the flame and achieving its tops in the front.

They stand out because the mix of the components which intervene in the

combustion is developed previous to the combustion. This is the way to control

its temperature. In these flames we can distinguish three zones. The first one,

called the pre-heat zone. There we can find low exothermic reactions, as the

components do not have enough temperature to react yet. Then, we can find the

reaction zone, it is the main part of the combustion. Once the components have

reached a concrete temperature, the combustion takes place. This is where the

heat is released. Finally, we can find the post-reaction zone, where the gases

from the combustion get cool.

The speed of this combustion can be modified by different factors, one of

the most important is the oxidant element and its proportion with respect to the

fuel. An increase in the proportion of oxygen will rise the combustion speed. The

initial temperature of the components is another factor. The higher the starting

temperature, the higher the growth in the flame speed. A larger range of flame

temperatures can be reached with this burning system and modifying those

parameters.


15

On the other hand, we can find the laminar flames. In these kinds of flames,

the mix of the components does not take place before the reaction. They mix at

the same time they are burning. Another characteristic to highlight is that the

temperature this system achieves is practically the stoichiometric.

This combustion system requires a complex study but one of the most

accurate description was made by Michael Faraday in 1908. He analysed the

flame of a candle, dividing the flame in three parts. The intern zone, composed

of the exhausted gases, which are gases which cannot get burned. Then, one

finds goes the intermedia zone, located in the limit where the fuel starts mixing

with surround oxygen, allowing its combustion. This is the region with higher

temperature so it emits light. The third region is composed by the external zone,

mainly integrated by oxygen. Here the free radicals formed in the areas of higher

temperature combine with oxygen, completing the oxidation or escaping in the

form of soot.

Figure 5 Some differences from premixed and non-premixed systems (Noble et al., 2021).

With independence of the kind of combustion, any burning system must

guarantee a certain margin of regulation. It is basically the maximum and

minimum power that the burner is able to produce before the appearance of

instabilities and incomplete combustion. The stability that each burner can

provide, understood as the ability to keep the combustion outside the designed

conditions should be assessed, as well as the polluting emissions it produces.


16

3.3 Performance and emissions

In this chapter, will be employed hydrogen with the combustion systems

exposed previously with the intention to get an appropriate balance between the

performances and the emissions from the power cycle. As has been said, in

diffusive flame combustors, the temperature of the flame nearly achieves

stoichiometric conditions, due to its high value it should be reduced by

incorporating some diluents, like nitrogen or steam. However, the addition of any

of those provokes a reduction into the performance and efficiency of the cycle,

as has been exposed previously.

On the other hand, we can find the premixed burners, they offer the

possibility to regulate the flame temperature variating the oxygen of the mix. This

option also involves its risks because realizing a stable premixed hydrogen flame

is not straightforward due to its high flame speed, it demands high air velocities

to obtain short mixing times and high turbulence rates. As another drawback,

premixed combustors may suffer from high pressure drops.

For this reason, gas turbine manufacturers are currently investigating and

developing different combustor geometries in order to obtain the same NOx

emissions and combustor pressure drops achieved in natural gas–fueled

combustors. The current industrial practice to burn H2 in a gas turbine consists of

employing diffusive flame combustors and prevent NOx formation by diluting the

fuel with steam or nitrogen which could be got from the steam cycle or an air

separation unit, respectively (Gazzani et al., 2014).

According to (Gazzani et al., 2014), the use of fuel dilution decreases the

efficiency due to different reasons: if nitrogen is used, it requires a compressor to

rise the pressure from the air separation unit (where we get the nitrogen from)

until the minimum combustor inlet pressure, this provokes an obvious rise into

the economic costs and a decrease into the net plant efficiency because this

compressor needs electricity to work. Should be mentioned that this discharge

pressure must be a little higher because combustors may require a fuel injection

pressure higher than the air pressure.

On the other hand, if we use steam and it is extracted from the heat

recovery of the cycle, we will also penalize the plant global efficiency due to the

energetic loss of mixing the steam with the fuel but also because of the reduction

of the temperature at the entrance of the turbine, necessary to deal with the

higher content of H2O in the fuel gases. Another disadvantage of the employment

of this diluent is the rise into the heat transfer coefficient, motivating an increase

into the temperature of the turbine blades. Therefore, the cooling system should


17

be modified to minimize this excess of heat or the temperature at the entrance

should be reduced (also reducing the outpower), but otherwise the blades of the

turbine may be damaged.

Must be also mentioned that every year the targets and regulations about

polluting emissions are becoming stricter, being really challenging to achieve with

the use of diluents in a diffusive flame.

Those reasons are why the actual industry is trying to develop premixed

hydrogen combustors. However, due to the special characteristics of the

hydrogen exposed before, this system of combustion should take into

consideration different issues: the larger flammability of the hydrogen and its

lower ignition temperatures, makes harder the task of mixing the air with the fuel

without provoking the autoignition, the higher flame speed provokes an unstable

combustion being able to produce a flashback which would damage the burners.

These burners may have a different geometry to limit the pressure drop with this

kind of combustion system is mandatory to achieve a perfect mixing of the air and

fuel before the combustion to control the emissions of NOx.

With the aim to compare the performances obtained by the different

systems of combustion, will be discussed the results from the simulation

developed by (Gazzani et al., 2014). In Table 3, we can see the heat balance and

performances of a gas turbine cycle used as reference to this simulation. This

cycle will be modified to be used with hydrogen as fuel.

This combined cycle-plant is composed by a single gas turbine, with triple

pressure level, a reheat heat recovery, a steam generator and one steam turbine.

Some assumptions have been adopted to allow the use of hydrogen like the

reduction of the temperature at the entrance of the turbine until its nominal

working value and the change into the compressor vanes to readjust the mass

flow (discussed in the first chapter of this research). Hydrogen is supposed to be

available at the required pressure.

Table 3 Performances and characteristics of the original power cycle used as reference, with NG as fuel. (Gazzani et al., 2014).


18

In the case of premixed combustion, the process steps do not require any

modification in comparation with the NG, the fuel is firstly preheated until 40ºC

and then it goes to the gas turbine. On the other hand, for the diffusive flame with

N2, the hydrogen is also preheated at 40ºC and combined with the nitrogen which

comes from a compressor, then the mix finds the air inside the combustion

chamber. Finally, for the steam solution, the hydrogen is preheated at the same

temperature and mixes with the steam, which comes at 300ºC, before the

combustion chamber.

According to the results obtained from the first simulation, which uses a

diffusive flame combustor with steam dilution,(Gazzani et al., 2014), we can

observe that for all the cases, working at nominal temperature at the entrance of

the turbine (with independence of the proportion of steam diluted), the gas turbine

cycle provides a larger power compared to the original NG, thanks to the

reduction into the compressor mass flow and its lower consume. In this case, the

cycle produces 325,4 MW approximately 15% higher than the in the case of NG.

Otherwise, the steam power plant decreases its efficiency in a range of

27% to 7%. The total power production of the combined plant keeps as practically

constant. Must be also highlighted that with those results we can observe a

tendency of reduction in the efficiency of the combined plant with the increasing

of the steam rate dilution in the hydrogen, at the same time as the steam dilution

rises, the cooling flow rises too, as commented before, due to the change into

heat transfer coefficient.

Looking at the cases with the same ratio of diluent but lower blade

temperature, can be concluded that the reduction of the temperature at the

entrance of the turbine reduces the turbine efficiency. To conclude, should be

noted that the volumetric heating value of the hydrogen mixture varies in the

range of 5800–9000 kJ/Nm3, which is fairly equivalent to syngas typically burned

in refinery plants.

Moving to the analysis of the second case, which uses a diffusive flame

combustor and nitrogen as diluent, we can observe a significant increase into the

power provided by the gas turbine, achieving 330 MW, nearly 17% more than the

reference cycle. In this case, the diluent does not affect the power provided by

the steam cycle. Otherwise, its power decreases a little bit compared to the

reference due to the lower mass flow along the circuit, which reduces its value in

22 kg/s. This reduction is motivated firstly, because the volumetric flow rate is

constant (about 166 m3/s), as the first stator geometry is already determined (it

has the rotor diameter and blade height).


19

Secondly, due to the specific volume of the working fluid increases. The

increase in the flue gas–specific heat (due to the higher H2O content) and the

cooling flow rates do not balance the previous effects, motivating the decrease

into the steam-power output (Gazzani et al., 2014).

In comparation with the previous case, the cooling flow required decreases

in a range from 2% to 8% but compared to the original natural gas cycle it has a

value of approximately 1.5% higher. The total plant efficiency reduces with the

increase into the nitrogen dilution (as happened with steam). The gas plant

efficiency decreases because the nitrogen requires the compressor before mixing

with the hydrogen. The volumetric heating value for this dilution is between the

range of 5000–8700 kJ/Nm3, lower than in the case of the steam.

The last simulation from the research (Gazzani et al., 2014) corresponds to

a premixed combustion, in this case we can observe a pressure drop of 3%

working at the nominal temperature. At the same time can be seen that increasing

the pressure drop in the combustion chamber provokes the reduction of the

power of the cycle and its efficiency, due to the enthalpy drop motivated. Other

factor involved is the reduction of the density of the gases at the entrance of the

turbine, reducing its density provokes a decrease into the cooling flow required

to protect the blades. Compared to the cycle of reference, we can observe an

increase in the efficiency of the combined plant.

In Table 4 can be observed a summary of some of the most relevant results.

In all cases, it is established a pressure ratio of 17, with the blades of the turbine

working in their nominal temperature.

Table 4 Summary of the main results obtained for each type of combustion.

Type of combustion

Diffusive flame and steam

Diffusive flame and nitrogen

Premixed flame

Mass along the compressor (kg/s)

638 626,8 540,7

Hydrogen mass flow (kg/s)

6,14 6,06 5,83

Diluent/ hydrogen 2,6 5,55 -

Power gas cycle (MW)

301,2 303,1 275,4

Power steam cycle (MW)

126,5 137,6 135,6

Total Power (MW) 424 420,2 407,2

Combined cycle net efficiency (%)

57,56 57,78 58,19


20

A stochiometric flame temperature of 2500K for both cases of diffusive

flame was determined and for the case of the premixed combustion a limited

pressure loss of 6,5%, data obtained from the simulation (Gazzani et al., 2014).

All data collected from the different simulations with different temperatures at the

entrance of the turbine and different ratio of diluent (in the case of nitrogen and

steam) can be consulted in table 5, 6 and 7 attached in the first appendix.

After analysing the data obtained, can be extracted some general

conclusions from this simulation, which are shown in Figure 6. In the chart, we

can observe the efficiency of each kind of combustion system discussed in

function of the stochiometric flame temperature, for the cases of diffusion flames

and in function of the pressure drop in the case of the premixed system.

We can clearly observe that in the case of the diffusion flame, the efficiency

provided when we use nitrogen as diluent is always higher than in the case of the

steam. This difference is even larger when we are working with lower

stochiometric flame temperatures (under 2400K), as can be seen comparing the

slope of both lines. Otherwise, when we move to flame temperatures higher than

2400K, the efficiency of both diluent trend the same value, around 58%. On the

other hand, can be concluded that the efficiency provided by the premixed

combustor system is higher than diffusive flame, with even a larger difference

when the pressure drop is lower.

Figure 6 Combined cycle electric efficiency for all the cases at nominal TIT: the efficiency for steam- and nitrogen-diluted combustors is plotted as function of the STFT (bottom x-axis), while the efficiency for the premixed combustor is plotted as function of the combustor relative pressure loss (upper x-axis).(Gazzani et al., 2014).


21

3.4 Premixed lean direct injection combustion

In view that premixed combustion provides the best efficiency, it is time to

deep in this system of combustion, showing their performances, efficiency and

characteristics.

Due to the high reactivity of the hydrogen those kinds of burners should take

into consideration the risk of flashback. Now will be discussed and assessed

some burners and techniques that could be employed with hydrogen.

With the aim to eliminate the risk of flashback, we can find the lean direct

injection (LDI), represented in Figure 7. LDI designs introduce fuel into excess

air at the combustion zone from numerous (often hundreds or thousands) of

locations as we can observe in

Figure 7, through the blue

canal, with the aim to achieve

rapid mixing. NOx emissions

are typically marginally higher

than a fully premixed system,

flame anchoring on the injector

may be an issue (York et al.,

2013). Micromixers are

composed by hundreds of

those small ducts used as

injectors as in the case of LDI

but they also include a

premixing zone before the

combustion, this mix can be

accomplished by several

methods like jet-in-crossflow mixing, coflow mixing in channels and swirl-based

mixing, among others.

Following these measures, the research (Marek et al., 2005) developed an

simulation with the aim to assess the emissions and efficiency of different burners

based on lean direct injection technology but with different configurations and

also compare them when using Jet-A, a fuel used to supply gas turbines engines

for planes. For every configuration with the aim to minimize the risks of

flashback will be employed short mixing times and high velocities.

A total of five different injectors technology were tested in this simulation,

one developed by NASA and the four reminders by some of the major fuel injector

manufacturers with several years of experience in this industry. Initially, the air

Figure7 Assembly of the configuration of a LDI burner (Marek et al., 2005).

7


22

pressure drop was limited at 4%. However, some burners breached it, was also

established an approximately air flow velocity at the entrance of the combustor of

30,5 m/s.

In the Figure 8 we can observe the different burners configuration tested in

the research (Marek et al., 2005) starting with the upper left, we can observe the

burner developed by NASA.

This injector contains two

opposing hydrogen jets in the

mixing tube, the jet penetration

and mixing system are based

on jet in crossflow technology.

According to (York et al.,

2013), this premixing system

uses the air flows through

numerous straight tubes of

millimetre-scale diameter in a

parallel array between a front

and back plate. The fuel

injection takes place through

many small holes distributed

along the tube walls. To

prevent flashbacks, the air

speed in these tubes must be really high. This multitube mixer achieves a low

pressure drop, not penalizing the efficiency of the combustion, as a result, it has

very short area of premixing. This air ducts have a diameter of 0,635 cm while

the hydrogen ducts will be smaller, being 0,508 cm.

Moving to the right, we can find the second burner tested, which will be

called as configuration 1. It is based on previously rocket injection technology

with a center “+” hydrogen jet and eight angled air jets mixing with the hydrogen.

For this concept, a rich region was created near the face for ignition and flame

holding using the four inner air jets of the air injection points. An immediate

quench section was then created just downstream using the four-remaining air

injection points. (Marek et al., 2005).

Continuum with the lower row, on the left part of the Figure 7, which will be

designed as configuration 2, we can find a burner similar to the first one but in

this case with triangular holes instead of circles. This configuration reduces the

size of the elements but otherwise increases its number combined with some

additional points to injects the hydrogen. These variations have the aim to

motivate the mix of both components to stimulate the combustion. Specifically, it

Figure 8 Pictures of four of the injectors tested (Marek et al., 2019).


23

contains 54 triangular LDI injectors with 3 hydrogen injection points per triangle

in a honeycomb pattern.

Finishing with the last image of Figure 7, configuration 3 is a conservative

design based on the already existing technology used in gas turbine. It is

composed by a single center hydrogen nozzle at the center of each hole with a

large amount of counter swirl to produce mixing. This design includes a single

hole hydrogen injection with counter swirl for each of the seven LDI holes.

The last configuration (configuration 4) tested is based on configuration

three (left lower row of Figure 7). However, the center hole is replaced by four

small radial diameter hydrogen jets per injection point, in this configuration does

not exist air swirl with the aim to reduce pressure drop. (Marek et al., 2005).

As can be observed in Figures 12 to 15 of the first appendix, where are

recorded the data obtained for each configuration comparing the use of hydrogen

and Jet-A as fuels, must be highlighted some issues. In configuration two, the

NOx levels were very low, achieving less than half of the Jet-A levels, but its

cooling and durability was compromised. It failed resulting in nonuniform mixing

and higher NOx (Marek et al., 2005).

In the case of the third configuration, the NOX emissions were half than in

the case of using Jet-A as fuel. The last configuration also proofed to be very

durable and able to be tested along a wide range of temperatures, showing an

emission of NOx much lower than when Jet-A is employed.

Figure 9 shows the emissions provided by each kind of configuration, in

every configuration based on lean direct injection is not observed high levels of

NOx emissions. As was predicted did not take place any phenomenon of

flashback or autoignition.

According to the data obtained, the configuration number four would be the

best option from the criteria of low emission and durability. As a general

phenomenon observed in this test (Marek et al., 2005) when there are the more

points of injection, the lower NOx emissions are produced. Must be also

mentioned that every injector exposed requires big effort in term of design and

manufacturing what implies an increase in its costs.


24

Figure 9 NOx emissions for all configurations, with a temperature at the entrance of 427 ºC (800ºF) and a pressure drop of 4% (Marek et al., 2005).


25

4.- Where could hydrogen be obtained from?

4.1 Classification of hydrogen

Once studied and discussed the effects and the changes which must be

accomplished to used hydrogen as fuel in heavy duty gas turbine and the different

system which can be used. This final part of the research will be mainly dedicated

to expose and discuss some of the methods we can use to get hydrogen to supply

these turbines, trying to show the benefits or disadvantages of each.

Before going through these methods, hydrogen can be classified depending

on the source it comes from and the technique employed, this classification is

based on colours. Firstly, we have the green hydrogen, it refers to the hydrogen

produced using renewable energy. It involves many systems like using wind,

solar or biomass energy to produce electricity and generate the electrolysis of the

water, dividing the molecules of H2O in, on the one hand H2 and on the other O2.

This is the most sustainable and friendliest with the environment way of producing

hydrogen because it does not produce any kind of greenhouse emissions to the

atmosphere. Unfortunately, it just involves 1% of the global production (Nogales,

2021).

In second place we can find the blue hydrogen, it involves the hydrogen

obtained from fossil sources, generally from natural gas or petroleum, this

process also involves the emission of CO2. In the case of blue hydrogen, this

greenhouse gases are captured so they do not reach the atmosphere. This

system is not as friendly with the environment as the first one but it does not

involve a significant damage to the atmosphere.

Continuing with the classification and really similar to the previous one, we

can find the gray hydrogen. It is hydrogen produced from the same sources as

blue hydrogen but in this case the greenhouse gases produced are not captured,

they are discharged into the atmosphere. This is the cheapest way of producing

hydrogen and nowadays corresponds to the 70% global production (Nogales,

2021). We can also find other colors like brown, it is the same as grey but using

coal instead of natural gas. Should also be mentioned the white hydrogen, it

refers to the hydrogen stored in the nature, in underground deposits. Probably

one day they could be extracted but currently we do not have this technology

available. Other lesser-known variants would be pink hydrogen, obtained through

nuclear energy using the electrolysis process, something that is also quite

sustainable (Nogales, 2021).


26

As can be seen in Figure 10, in 2009 approximately 96% of the global

production of hydrogen employed fossil resources, generation the emission of

greenhouse gases, mostly of those from the reforming of methane (system

discussed in the next subchapter).

Figure 10 Hydrogen production processes (blue shares correspond to those methods that use fossil fuels (Balat & Balat, 2009).

4.2 Methods to produce hydrogen

Following with this research will be briefly explained three ways of producing

hydrogen, the first one through the electrolysis of water, then by steam reforming

of natural gas and finally by the gasification of coal.

The first method exposed will be the electrolysis of water, it is mainly used

when the amount of hydrogen required is not much large. Nowadays, it is the

best-known system to produce

hydrogen, it is based on the

separation of the atoms of water

using electricity.

The electrodes, cathode and

anode, are situated in the solution

and generate the movement of

electrons. The hydrogen is formed

in the cathode, at the same time

but in half of the volume of the

hydrogen, due to the composition

of the molecule of water, the

oxygen does it in the anode. With Figure 11 Outline of an electrolitic cell (Angelovska, 2016)


27

the aim to improve the production of both elements it is usually changed the

composition of the water, by the addition of salts, to improve the speed of the

reaction, according to the research (Fernandez, 2005), the efficiency of this

system is around 75%.

This hydrogen produced must be purify because it contains oxygen

impurities and a certain level of humidity. The costs of production by this method

are around 4,9-5,6 KWh per m3 of hydrogen produced, resulting two times more

expensive than reforming natural gas. (G.Fierro, 2011). With the intention to

reduce these costs, we can also find the electrolysis but in vapor phase. In this

case the costs of production are proportional to the electricity used and it

decreases with the temperature, at 1.500K the costs required to produce H2 are

approximately 50% lower (G.Fierro, 2011).

As has been shown, the requirement for this system is the use of electricity

to divide the molecules of water. Thus, depending on the source, we are getting

this electricity from we could be talking of a completely sustainable production,

with no pollution of the environment or otherwise, of a production system not such

respectful with the environment.

The second technique to the production of hydrogen is based on the steam

reforming of natural gas, specially of the methane it contains. This method relates

the two components discussed along this research, below is shown the

stochiometric reaction which takes place:

𝐶𝐻4 + 𝐻2𝑂 — > 𝐶𝑂 + 3𝐻2 (3)

As can be seen, the NG reacts with the steam to produce carbon monoxide

and hydrogen. This reaction develops at high temperature and pressure,

previously the NG requires a process of purification to remove sulfur impurities it

may contain. Then the stream of clean methane reacts in a reactor with the

presence of nickel, used as catalysator, to increase the speed of the reaction.

The exhausted gases are rich in hydrogen but with a large amount of carbon

monoxide. These gases react in another reactor with the presence of steam to

produce more hydrogen. The resulting gas has a high proportion of hydrogen,

with CO2 and lower quantity of non-transformed methane and carbon monoxide

in around 1% of the volume (G.Fierro, 2011).

The modern plants also incorporate units to purify the gases, obtaining a

99,999% pure hydrogen in volume (G.Fierro, 2011). According to (Fernandez,


28

2005), the efficiency of this system is around 70%, a little lower than in the case

of electrolysis.

This process is widely used in the industry because is the cheapest,

according to the “Office of Energy efficiency & renewable energy”, the 95% of the

hydrogen produced in the United Stated is produced with this system.

Instead of methane could also be used methanol which reacts with steam

to produce hydrogen. This is an endothermic reaction which uses the heat coming

from burning some part of the methanol. As in the case of the methane, the

exhausted gases must also be purified. On the other hand, this reaction is simpler

because does not take into account the formation of intermediate oxygenates

(G.Fierro, 2011), but due to its higher cost it is less used.

The third and last system which will be discussed consist into gasification of

coal, it consists in the production of hydrogen from coal. It starts by heating the

solid coal until it becomes a gas, then this gas reacts with steam and oxygen to

produce hydrogen, carbon monoxide and carbon dioxide. Deepening a little into the reaction, this process requires two reactions, in

the first one we obtain carbon monoxide from the coal and the second one

transforms this monoxide in carbon dioxide, both give off hydrogen, each one

takes place in a different reactor because they requiere different temperatures.

This system is nearly two times more expensive than obtaining hydrogen

from NG, due to the ratio of hydrogen-carbon in the NG is 4:1 while in coal is

0,8:1 (Fernandez, 2005), and must also be highlighted that the emissions of

carbon dioxide it produces are larger than in case of the NG so they should be

captured.

Once exposed these different systems of producing hydrogen, must be said

that one of the most important reason of moving from fossil fuels to hydrogen is

because of the reduction of the environmental impact. This change would not be

so useful if the production of this hydrogen keeps polluting the atmosphere. This

is an issue which requires more investigation and development with the aim to

create new and affordable systems or methods which employ renewable sources

to carry out this task.


29

5.- Conclusion

Currently, a world without electricity would be unthinkable. One part of this

electricity is produced using heavy duty gas turbines. With the aim to respect the

environment, they could be transformed into low or zero-carbon emitting systems.

This task can be done through pre- or post-combustion options. Pre-combustion

choices include the use of hydrogen or any other renewable source. Post-

combustions options include carbon capture and oxyfuels (Goldmeer, 2020). This

investigation opts for the first solution.

During this research, the implications and the requirements of changing

from a NG gas turbine to one fueled with hydrogen have been shown and some

conclusions were discussed. Regarding the process of combustion, we can find

two different systems: diffusive and premixed flame. The first one is widely used

in this kind of turbines, while the second one is under development.

Due to properties of hydrogen, specially its higher stochiometric flame

temperature compared to NG, when using a diffusive combustion system, a

diluent should be included, otherwise the production of NOx would be excessive.

We can find two diluents: nitrogen and steam. According to the different

experiments and research reviewed, it can be concluded that nitrogen provides

the best performances, generating the lower efficiency drop.

To accommodate this new diluent and its impact on the turbomachinery,

different methods have been discussed. Overall, adjusting the variable guide

vanes of the compressor appears to be a better solution than changing the

pressure ratio. However, the best solution would consist in the design and

construction of a new cycle, but this is not the aim of this research.

Moving to the premixed system, it has proved to be the one with higher

efficiency in different tests. However, this system still requires development and

research even though many advances have been made. This system has to

mainly take care of the risk of flashback. With the aim to minimize them, the lean

direct injection has been explained in this research. They are a system of

combustors which operate in a fuel lean mode and achieve rapid mixing of fuel

and air using hundreds or thousands of small-scale mixers. They provide a good

efficiency with a low production of NOx, although unfortunately, they require a

hard and difficult process of design, manufacture and production, circumstances

which increase its price.


30

However, one of the most limiting factors regarding hydrogen is its

production. Nowadays, established, large-scale, renewable-based hydrogen

production does not exist. In this research, three different systems have been

explained, but it is mandatory to find a non-pollutant system of production which

allows to produce hydrogen in large scale to supply these turbines.

As a general conclusion, it can be said that pre-existing gas turbine cycles

can be run with hydrogen. This will imply a huge step to achieve the reduction of

greenhouse gases trying to reverse the current climate change that we are facing

and moving towards a more sustainable world.

Figures and tables

31

Figures and tables

Tables

Table Description Page

1 Comparation of properties between hydrogen and methane. 6

2 Comparation of the results for each of the solutions implemented. 10

3 Performances and characteristics of the original power cycle used as reference, with NG as fuel.

17

4 Summary of the main results obtained for each type of combustion. 19

5 Results for the hydrogen-fueled combined cycle with diffusive flame combustor and steam dilution. Nomenclature: H2O: steam dilution; NOx1-2-3-4: ordered with decreasing STFT (2575, 2500, 2350, 2200 K); TIT1-2-3: ordered with decreasing TIT (or blade metal temperature), 1 is the nominal temperature, 2 and 3 correspond to a metal temperature decrease of 20ºC or 40ºC, respectively.

35

6 Results for the hydrogen-fueled combined cycle with diffusive flame combustor and nitrogen dilution. Nomenclature: N2: nitrogen dilution; NOx1-2-3-4: ordered with decreasing STFT (2575, 2500, 2350, 2200 K); TIT1-2-3: ordered with decreasing TIT (or metal blade temperature), 1 is the nominal temperature, 2 and 3 correspond to a decrease of 20ºC or 40ºC on the metal, respectively.

35

7 Results for the hydrogen-fueled combined cycle with premixed combustor. Nomenclature: prem: premixed combustor; Dp 1-2-3: ordered with increasing combustor pressure loss (3.0%, 6.5%, 10.0%); TIT1-2-3: ordered with decreasing TIT (or metal blade temperature), 1 is the nominal temperature, 2 and 3 correspond to a decrease of 20ºC or 40ºC on the metal, respectively.

36

Figures

Figure Description Page

1 Global primary energy consumption by source. 1

2 Representation of a Brayton cycle with its more significant points. 2

3 Variation of the stoichiometric flame temperature and of the inlet volume flow rate and isentropic enthalpy drop of a hydrogen and steam fueled gas turbine with respect to the reference natural gas case.

7

4 Variation of the stoichiometric flame temperature and of the inlet volume flow rate and isentropic enthalpy drop of a hydrogen and nitrogen fueled gas turbine with respect to the reference natural gas case.

8

5 Some differences from premixed and non-premixed systems. 15

6 Combined cycle electric efficiency for all the cases at nominal TIT: the efficiency for steam- and nitrogen-diluted combustors is plotted as function of the STFT (bottom x-axis), while the efficiency for the premixed combustor is plotted as function of the combustor relative pressure loss (upper x-axis).

20

7 Assembly of the configuration of a LDI burner. 21

8 Pictures of four of the injectors tested. 22

9 NOx emissions for all configurations, with a temperature at the entrance of 426,667 ºC (800ºF) and a pressure drop of 4%.

24

10 Hydrogen production processes (blue shares correspond to those methods that use fossil fuels.

26

Figures and tables

32

11 Outline of an electrolitic cell. 26

12 Emissions of the injector designed by NASA, with 63,5 mm of liner, an inlet temperature of 427ºC and an air flow speed of 40 m/s.

36

13 Emissions of the injector in configuration 1, with 63,5 mm of liner, an inlet temperature of 443,3ºC and an air flow speed of 30,5 m/s.

37


37


38

Bibliography

33

Bibliography

Angelovska, E. (2016). Difference Between Galvanic Cells and Electrolytic Cells.

http://www.differencebetween.net/science/difference-between-galvanic-

cells-and-electrolytic-cells/.

ARREGLE, J., Broatch, J. A., Galindo, J., Luján, J. M., Pastor, J. V., Payri, R.,

Serrano, J. R., & Torregrosa, A. J. (2002). Procesos y tecnología de

máquinas y motores térmicos. Universidad Politécnica De Valencia, 1(1),

399. http://books.google.com/books?id=DExo1KyDnckC&pgis=1

Badía, C. F. (2005). 2.4 Propiedades del Hidrógeno. Tesis Doctoral, Química

Orgánica, Universidad de Sevilla, España, 73–81.

Balat, M., & Balat, M. (2009). Political, economic and environmental impacts of

biomass-based hydrogen. International Journal of Hydrogen Energy, 34(9),

3589–3603. https://doi.org/10.1016/j.ijhydene.2009.02.067

Bannister, R. L., Newby, R. A., & Yang, W. C. (1997). Development of a

hydrogen-fueled combustion turbine cycle for power generation.

Proceedings of the ASME Turbo Expo, 2(January).

https://doi.org/10.1115/97-GT-014

Chiesa, P., Lozza, G., & Mazzocchi, L. (2005). Using hydrogen as gas turbine

fuel. Journal of Engineering for Gas Turbines and Power, 127(1), 73–80.

https://doi.org/10.1115/1.1787513

Díaz, M., & Ascencio, D. (2009). Análisis De Las Emisiones De Gases

Contaminantes De Un Ciclo Brayton Como Función De La Eficiencia

Exergética Global. 1–5.

Fadok, J., & Diakunchak, I. (2008). Advanced hydrogen turbine development

update. 25th Annual International Pittsburgh Coal Conference, PCC -

Proceedings, September 2007.

Fernandez, C. (2005). Sistema de Energía del Hidrogeno. Energética Del

Hidrogeno. Contexto, Estado AActual y Perspectivas de Futuro, 91–126.

Filn, D. (2019). Working towards 100% hydrogen. Siemens Report, 2, 9–13.

Gazzani, M., Chiesa, P., Martelli, E., Sigali, S., & Brunetti, I. (2014). Using

hydrogen as gas turbine fuel: Premixed versus diffusive flame combustors.

Journal of Engineering for Gas Turbines and Power, 136(5).

https://doi.org/10.1115/1.4026085

Bibliography

34

G.Fierro, J. L. (2011). Ways to produce hydrogen. Instituto de Catálisis y

Petroleoquímica (CSIC).

http://www.fgcsic.es/lychnos/es_es/articulos/hidrogeno_metodologias_de_

produccion

Goldmeer, J. (2020). Solving the challenge of lean hydrogen premix combustion

with highly reactive fuels.

https://www.turbomachinerymag.com/view/solving-the-challenge-of-lean-

hydrogen-premix-combustion-with-highly-reactive-fuels

Gutierrez Jodral, L. (1968). ELHIDRÓGENO, COMBUSTIBLE DELFUTURO

LUIS. Hawaii Medical Journal, 27(5), 468–469.

Iea Org. (2021). Data about energy. Journal of Chemical Information and

Modeling. https://www.iea.org/countries

Koç, Y., Yağlı, H., Görgülü, A., & Koç, A. (2020). Analysing the performance, fuel

cost and emission parameters of the 50 MW simple and recuperative gas

turbine cycles using natural gas and hydrogen as fuel. International Journal

of Hydrogen Energy, 45(41), 22138–22147.

https://doi.org/10.1016/j.ijhydene.2020.05.267

Marek, C. J., Smith, T. D., & Kundu, K. (2005). Using Lean Direct Injection. Joint

Propulsion Conference and Exhibit Tucson., AIAA–2005–, 1–27.

Noble, D., Wu, D., Emerson, B., Sheppard, S., Lieuwen, T., & Angello, L. (2021).

Assessment of Current Capabilities and Near-Term Availability of Hydrogen-

Fired Gas Turbines Considering a Low-Carbon Future. Journal of

Engineering for Gas Turbines and Power, 143(4), 1–10.

https://doi.org/10.1115/1.4049346

Nogales, M. (2021). Kind of hydrogen based on origen.

https://noticias.coches.com/consejos/tipos-de-hidrogeno-por-su-

obtencion/421373

UnionGas. (2017). Components of the natural gas.

https://www.uniongas.com/about-us/about-natural-gas/chemical-

composition-of-natural-gas

York, W. D., Ziminsky, W. S., & Yilmaz, E. (2013). Development and testing of a

low NOx hydrogen combustion system for heavy-duty gas turbines. Journal

of Engineering for Gas Turbines and Power, 135(2), 1–8.

https://doi.org/10.1115/1.4007733

https://www.uniongas.com/about-us/about-natural-gas/chemical-composition-of-natural-gas

https://www.uniongas.com/about-us/about-natural-gas/chemical-composition-of-natural-gas

First Appendix

35

Appendix

Table 5 Results for the hydrogen-fueled combined cycle with diffusive flame combustor and steam dilution. Nomenclature: H2O: steam dilution; NOx1-2-3-4: ordered with decreasing STFT (2575, 2500, 2350, 2200 K); TIT1-2-3: ordered with decreasing TIT (or blade metal temperature), 1 is the nominal temperature, 2 and 3 correspond to a metal temperature decrease of 20ºC or 40ºC, respectively (Gazzani et al., 2014).

Table 1 Results for the hydrogen-fueled combined cycle with diffusive flame combustor and steam dilution. Nomenclature: H2O: steam dilution; NOx1-2-3-4: ordered with decreasing STFT (2575, 2500, 2350, 2200 K); TIT1-2-3: ordered with decreasing TIT (or blade metal temperature), 1 is the nominal temperature, 2 and 3 correspond to a metal temperature decrease of 20ºC or 40ºC, respectively.(Gazzani et al., 2014).

Table 6 Results for the hydrogen-fueled combined cycle with diffusive flame combustor and nitrogen dilution. Nomenclature: N2: nitrogen dilution; NOx1-2-3-4: ordered with decreasing STFT (2575, 2500, 2350, 2200 K); TIT1-2-3: ordered with decreasing TIT (or metal blade temperature), 1 is the nominal temperature, 2 and 3 correspond to a decrease of 20ºC or 40ºC on the metal, respectively (Gazzani et al., 2014).

Table 2 Results for the hydrogen-fueled combined cycle with diffusive flame combustor and nitrogen dilution. Nomenclature: N2: nitrogen dilution; NOx1-2-3-4: ordered with decreasing STFT (2575, 2500, 2350, 2200 K); TIT1-2-3: ordered with decreasing TIT (or metal blade temperature), 1 is the nominal temperature, 2 and 3 correspond to a decrease of 20ºC or 40ºC on the metal, respectively.(Gazzani et al., 2014).

First Appendix

36

Table 7 Results for the hydrogen-fueled combined cycle with premixed combustor. Nomenclature: prem: premixed combustor; Dp 1-2-3: ordered with increasing combustor pressure loss (3.0%, 6.5%, 10.0%); TIT1-2-3: ordered with decreasing TIT (or metal blade temperature), 1 is the nominal temperature, 2 and 3 correspond to a decrease of 20ºC or 40ºC on the metal, respectively (Gazzani et al., 2014).

Figure 12 Emissions of the injector designed by NASA, with 63,5 mm of liner, an inlet temperature of 427ºC and an air flow speed of 40 m/s(Marek et al., 2005).

First Appendix

37

Figure 13 Emissions of the injector in configuration 1, with 63,5 mm of liner, an inlet temperature of 443ºC and an air flow speed of 30,5 m/s (Marek et al., 2019).

Figure 14 Emissions of the injector in configuration 3, with 63,5 mm of liner, an inlet temperature of 427ºC and an air flow speed of 21 m/s (Marek et al., 2019).

First Appendix

38

Figure 15 Emissions of the injector in configuration 4, with 88,9 mm of liner, an inlet temperature of 427ºC and an air flow speed of 18 m/s (Marek et al., 2019).

study of a gas turbine cycle with hydrogen combustion

Documents